JP7544883B2 - Interaction Between LUT and AMVP - Google Patents

Description

CROSS-REFERENCE TO RELATED APPLICATIONS Under applicable patent laws and/or regulations under the Paris Convention, this application is intended to timely claim priority to and the benefit of International Patent Application No. PCT/CN2018/093663, filed June 29, 2018, International Patent Application No. PCT/CN2018/105193, filed September 12, 2018, and International Patent Application No. PCT/CN2019/072058, filed January 16, 2019. Under the laws of the United States, for all purposes, the entire disclosures of International Patent Application No. PCT/CN2018/093663, International Patent Application No. PCT/CN2018/105193, and International Patent Application No. PCT/CN2019/072058 are incorporated by reference as part of the disclosure of this application.

This patent specification relates to video encoding and decoding techniques, devices and systems.

Despite advances in video compression, digital video still accounts for the largest bandwidth usage on the Internet and other digital communications networks. As the number of connected user devices capable of receiving and displaying video increases, the bandwidth demands for digital video use are expected to continue to grow.

This specification discloses methods, systems, and devices for encoding and decoding digital video.

In one exemplary aspect, a method of video decoding is provided that includes maintaining tables, each table including a set of motion candidates, each motion candidate being associated with corresponding motion information, and converting between a first video block and a bitstream representation of a video including the first video block, the converting including processing the motion information of the first video block using at least a portion of the set of motion candidates as predictors.

In yet another representative aspect, the various techniques described herein may be implemented as a computer program product stored on a non-transitory computer-readable medium. The computer program product includes program code for performing the methods described herein.

Details of one or more implementations are set forth in the accompanying attachments, drawings, and description below. Other features will be apparent from the description and drawings, and from the claims.

FIG. 2 is a block diagram illustrating an example implementation of a video encoder. 1 shows a macroblock division in the H.264 video coding standard. An example of dividing a coding block (CB) into prediction blocks (PU) will be shown. 1 illustrates an example implementation for subdividing a CTB into CBs and Transform Blocks (TBs), with solid lines indicating CB boundaries and dotted lines indicating TB boundaries, including an example CTB with its division, and the corresponding quadtree. 1 shows an example of a quad tree binary tree (QTBT) structure for dividing video data. 1 shows an example of video block division. An example of a quadtree division is shown below. 1 shows an example of a tree-type signaling. 13 illustrates an example of a derivation process for building a merge candidate list. 1 shows examples of spatial merge candidate locations. 1 shows examples of candidate pairs considered for redundancy check of spatial merge candidates. 4 shows examples of the location of the second PU for N×2N and 2N×N partitions. 13 illustrates the scaling of motion vectors for temporal merge candidates. Shows the candidate locations of temporal merge candidates and their co-located pictures. 1 illustrates an example of a combined bi-predictive merge candidate. 13 shows an example of a process for deriving motion vector prediction candidates. 13 illustrates an example of motion vector scaling for spatial motion vector candidates. 1 illustrates an exemplary Alternative Temporal Motion Vector Prediction (ATMVP) for motion prediction of a CU. 1 illustrates a pictorial example of source block and source picture identification. An example of one CU with four sub-blocks and neighboring blocks is shown. 1 shows an example of bilateral matching. An example of template matching will be given below. An example of unilateral motion estimation (ME) in Frame Rate Up Conversion (FRUC) is shown. 1 shows an example of DMVR based on bilateral template matching. 1 shows an example of spatially neighboring blocks used to derive spatial merging candidates. 13 shows an example of how to select a representative location for updating a lookup table. An example of updating a lookup table with a new set of motion information is given. An example of updating a lookup table with a new set of motion information is given. FIG. 2 is a block diagram illustrating an example of a hardware platform for implementing the visual media decoding or visual media encoding techniques described herein. 4 is a flow chart of another exemplary method of video bitstream processing. 1 shows an example of a decoding flowchart according to the proposed HMVP method. An example of updating a table using the proposed HMVP method is shown. 1 shows an example of a redundancy elimination based LUT update method (eliminating one redundant motion candidate). 1 shows an example of a redundancy elimination based LUT update method (eliminating one redundant motion candidate). 1 shows an example of a redundancy elimination based LUT update method (eliminating multiple redundant motion candidates). 1 shows an example of a redundancy elimination based LUT update method (eliminating multiple redundant motion candidates). An example of the difference between type 1 blocks and type 2 blocks is shown below.

To improve video compression rates, researchers are constantly seeking new techniques for encoding video.

1. Introduction

This specification relates to video coding technology. Specifically, it relates to coding of motion information in video coding (e.g., merge mode, AMVP mode). It may be applied to existing video coding standards such as HEVC, or may be applied to finalize a standard (Versatile Video Coding). The present invention is also applicable to future video coding standards or video codecs.

Brief description

Video coding standards have evolved primarily through the development of well-known ITU-T and ISO/IEC standards. ITU-T produced H.261 and H.263, ISO/IEC produced MPEG-1 and MPEG-4 Visual, and the two organizations jointly produced H.262/MPEG-2 Video, H.264/MPEG-4 AVC (Advanced Video Coding), and H.265/HEVC standards. Since H.262, video coding standards have been based on a hybrid video coding structure in which temporal prediction and transform coding are utilized. An example of a typical HEVC encoder framework is shown in Figure 1.

2.1 Partition structure

2.1.1 Partition tree structure in H.264/AVC

The core of the coding layer in previous standards was a macroblock containing a 16x16 block of luma samples and, in the case of normal 4:2:0 color sampling, two corresponding 8x8 blocks of chroma samples.

Intra-coded blocks use spatial prediction to exploit spatial correlation between pixels. We define two partitions: 16x16 and 4x4.

Inter-coded blocks use temporal prediction instead of spatial prediction by estimating the motion between pictures. Motion can be estimated independently for a 16x16 macroblock or any of its sub-macroblock partitions: 16x8, 8x16, 8x8, 8x4, 4x8, 4x4 (see Figure 2). Only one motion vector (MV) per sub-macroblock partition is allowed.

2.1.2 Partition tree structure in HEVC

In HEVC, CTUs are partitioned into CUs using a quad-tree structure called a coding tree to accommodate various local features. The decision to code a picture region using inter-picture (temporal) or intra-picture (spatial) prediction is made at the CU level. Each CU can be further partitioned into one, two or four PUs depending on the PU partition type. Inside one PU, the same prediction process is applied and related information is sent to the decoder on a PU-by-PU basis. After applying the prediction process based on the PU partition type to obtain the residual block, the CU can be partitioned into transform units (TUs) based on another quad-tree structure similar to the coding tree for CUs. One of the key features of the HEVC structure is that it has a multiple partition concept including CU, PU and TU.

Below we highlight various features related to hybrid video coding using HEVC.

1) Coding Tree Unit and Coding Tree Block (CTB) Structure. A similar structure in HEVC is the coding tree unit (CTU), which has a size selected by the encoder and may be larger than a conventional macroblock. A CTU consists of a luma CTB and a corresponding chroma CTB and syntax elements. The size LxL of the luma CTB can be selected as L=16, 32, or 64 samples, with larger sizes generally enabling better compression. HEVC then supports splitting the CTB into smaller blocks using a tree structure and quadtree-like signaling.

2) Coding Units (CUs) and Coding Blocks (CBs): The quadtree syntax of a CTU specifies the size and location of its luma and chroma CBs. The root of the quadtree is associated with the CTU. The size of the luma CTB is therefore the maximum size supported for the luma CB. The partitioning of a CTU into luma and chroma CBs is signaled together. One luma CB and usually two chroma CBs, together with the associated syntax, form one coding unit (CU). A CTB may contain only one CU or may be partitioned to form multiple CUs, each with its associated partition into prediction units (PUs) and one tree of transform units (TUs).

3) Prediction Units and Prediction Blocks (PBs): The decision of whether to code a picture region using inter-picture or intra-picture prediction is made at the CU level. The partitioning structure of the PU has its root at the CU level. Based on the decision of the basic prediction type, the size of the luma and chroma CBs can then be further partitioned and predicted from the luma and chroma prediction blocks (PBs). HEVC supports variable PB size samples from 64x64 to 4x4. Figure 3 shows an example of allowed PBs for an MxM CU.

4) TUs and Transform Blocks: The prediction residual is coded using a block transform. The TU tree structure has its root at the CU level. This luma CB residual may be identical to the luma transform block (TB) or may be further split into smaller luma TBs. Similarly for the chroma TB. For square TB sizes 4x4, 8x8, 16x16, and 32x32, integer basis functions similar to those of the discrete cosine transform (DCT) are specified. For the 4x4 transform of the luma intra-picture prediction residual, an integer transform derived from a form of the discrete sine transform (DST) is alternatively specified.

Figure 4 shows an example of subdivision of a CTB into CBs [and transform blocks (TBs)]. Solid lines indicate CB boundaries, dotted lines indicate TB boundaries. (a) CTB and its division (b) corresponding quadtree.

2.1.2.1 Dividing the tree structure into transformation blocks and units

For residual coding, the CB can be recursively split into transform blocks (TBs). This split is signaled by the residual quadtree. We only specify square CB and TB splits so that one block can be recursively split into quadrants as shown in Figure 4. For a given luma CB of size MxM, a flag signals whether it is split into four blocks of size M/2xM/2. If further splits are possible, each quadrant is assigned a flag indicating whether it is split into four quadrants, as signaled by the maximum depth of the residual quadtree indicated in the SPS. The leaf node blocks resulting from the residual quadtree are the transform blocks that are further processed by transform coding. The encoder indicates the maximum and minimum luma TB sizes it will use. If the CB size is larger than the maximum TB size, the split is done implicitly. If the split would result in a luma TB size smaller than the indicated minimum, no split is done implicitly. The chroma TB size is half the luma TB size in each dimension, except when the luma TB size is 4x4, in which case one 4x4 chroma TB is used for the area covered by four 4x4 luma TBs. For intra-picture predicted CUs, decoded samples of the nearest neighboring TBs (inside or outside the CB) are used as reference data for intra-picture prediction.

In contrast to previous standards, the HEVC design allows a TB to span multiple PBs for inter-picture predicted CUs, maximizing the potential coding efficiency benefits of quadtree TB partitioning.

2.1.2.2 Parent and child nodes

The CTB is divided based on a quadtree structure, whose nodes are coding units. The multiple nodes in the quadtree structure include leaf nodes and non-leaf nodes. The leaf nodes do not have child nodes in the tree structure (i.e., the leaf nodes are not further divided). The non-leaf nodes include a root node of the tree structure. The root node corresponds to a first video block (e.g., the CTB) of the video data. In each non-root node of the multiple nodes, the non-root node corresponds to a video block that is a sub-block of a video block corresponding to a parent node in the tree structure of the respective non-root node. Each non-leaf node of the multiple non-leaf nodes has one or more child nodes in the tree structure.

2.1.3 Quadtree + binary tree block structure with larger CTU in JEM

To explore future video coding technologies beyond HEVC, VCEG and MPEG jointly established the Joint Video Exploration Team (JVET) in 2015. Since then, many new methods have been adopted by JVET and incorporated into reference software called Joint Exploration Mode (JEM).

2.1.3.1 Division structure of QTBT block

Unlike HEVC, the QTBT structure removes the concept of multiple partition types, i.e., it removes the separation of the concepts of CU, PU, and TU, and improves the flexibility of the shape of CU partitions. In the QTBT block structure, a CU can have either a square or a rectangle. As shown in Figure 5, first, a coding tree unit (CTU) is divided by a quadtree structure. The leaf nodes of the quadtree are further divided by a binary tree structure. There are two types of division in the binary tree: symmetric horizontal division and symmetric vertical division. The leaf nodes of the binary tree are called coding units (CUs), and this segmentation is used for prediction and transform processing without further division. This means that in the coded block structure of QTBT, CUs, PUs, and TUs have the same block size. In JEM, a CU often consists of coding blocks (CBs) of different color components, e.g., for P and B slices in 4:2:0 chroma format, one CU contains one luma CB and two chroma CBs, and a CU often consists of CBs of a single component, e.g., for I slices, one CU contains only one luma CB or only two chroma CBs.

The following parameters are specified for the QTBT partitioning scheme:
-CTU size: size of root node of one quadtree, same concept as HEVC -MinQTSize: minimum allowable quadtree leaf node size -MaxBTSize: maximum allowable binary tree root node size -MaxBTDepth: maximum allowable binary tree depth -MinBTSize: minimum allowable binary tree leaf node size

In one example of a QTBT partitioning structure, the size of the CTU is set as 128x128 luma samples with two corresponding 64x64 blocks of chroma samples, MinQTSize is set as 16x16, MaxBTSize is set as 64x64, MinBTSize (for both width and height) is set as 4x4, and MaxBTDepth is set as 4. A quadtree partition is first applied to the CTU to generate quadtree leaf nodes. The quadtree leaf nodes can have sizes from 16x16 (i.e., MinQTSize) to 128x128 (i.e., CTU size). If a leaf quadtree node is 128x128, it will not be further partitioned by the bipartite tree since its size exceeds MaxBTSize (i.e., 64x64). Otherwise, the leaf quadtree node may be further split by the binary tree. Thus, the leaf node of this quadtree is also the root node of the binary tree, whose depth is 0. If the depth of the binary tree reaches MaxBTDepth (i.e., 4), no further splits are considered. If the width of a binary tree node is equal to MinBTSize (i.e., 4), no further horizontal splits are considered. Similarly, if the height of a binary tree node is equal to MinBTSize, no further vertical splits are considered. The leaf node of the binary tree is further processed by the prediction and transform process without further splits. In JEM, the maximum CTU size is 256x256 luma samples.

Figure 5 (left) shows an example of block partitioning using QTBT, and Figure 5 (right) shows the corresponding tree representation. Solid lines represent quadtree partitioning, and dotted lines represent binary tree partitioning. At each partition (i.e., non-leaf) node of the binary tree, a flag is signaled to indicate which partition type (i.e., horizontal or vertical) is used, where 0 represents a horizontal partition and 1 represents a vertical partition. In the case of quadtree partitioning, there is no need to indicate the partition type, since a quadtree partition always partitions a block horizontally and vertically, generating four sub-blocks of equal size.

Furthermore, the QTBT scheme supports the ability for luma and chroma to have separate QTBT structures. Currently, for P and B slices, the luma and chroma CTBs in one CTU share the same QTBT structure. However, for I slices, the luma CTBs are split into CUs by a QTBT structure, and the chroma CTBs are split into chroma CUs by another QTBT structure. This means that one CU in one I slice consists of one coded block of one luma component or one coded block of two chroma components, and one CU in one P or B slice consists of coded blocks of all three color components.

In HEVC, inter prediction for small blocks is restricted to reduce memory accesses for motion compensation, resulting in no bi-prediction being supported for 4x8 and 8x4 blocks, and no inter prediction being supported for 4x4 blocks. In JEM's QTBT, these restrictions are removed.

2.1.4 VVC ternary tree

In some embodiments, tree types other than quadtrees and binary trees are supported. In this implementation, we introduce two or more ternary tree (TT) partitions, i.e., horizontal and vertical center-side ternary trees, as shown in Figures 6(d) and (e).

Figure 6 shows (a) quadtree division, (b) vertical binary tree division, (c) horizontal binary tree division, (d) vertical center side ternary tree division, and (e) horizontal center side ternary tree division.

In some implementations, there are two levels of trees: region tree (quadtree) and prediction tree (binary or ternary). CTUs are first split by region tree (RT). RT leaves may be further split by prediction tree (PT). PT leaves may also be further split by PT until a maximum PT depth is reached. PT leaves are the basic coding units. For convenience, we still refer to them as CUs. A CU cannot be further split. Both prediction and transformation are applied to CUs, similar to JEM. We refer to the whole partition structure as a "multi-type tree".

2.1.5 Split structure

The tree structure used in this response is called the Multi-Tree Type (MTT), which is a generalization of QTBT. In QTBT, as shown in Figure 5, the coding tree unit (CTU) is first divided into a quadtree structure. The leaf nodes of the quadtree are further divided into binary trees.

The basic structure of MTT consists of two types of tree nodes: region tree (RT) and prediction tree (PT), which support nine types of partitions, as shown in Figure 7.

Figure 7 shows (a) quadtree partitioning, (b) vertical binary tree partitioning, (c) horizontal binary tree partitioning, (d) vertical ternary tree partitioning, (e) horizontal ternary tree partitioning, (f) horizontal upper asymmetric binary tree partitioning, (g) horizontal lower asymmetric binary tree partitioning, (h) vertical left asymmetric binary tree partitioning, and (i) vertical right asymmetric binary tree partitioning.

A region tree can recursively split a CTU into square blocks to become leaf nodes of the region tree of size 4x4. At each node in the region tree, a prediction tree can be formed from one of three tree types: binary tree (BT), ternary tree (TT), and asymmetric binary tree (ABT). In PT partitioning, it is forbidden to have quad tree partitions in the branches of the prediction tree. As in JEM, the luma and chroma trees are divided into I slices. The signaling method of RT and PT is shown in Figure 8.

2.2 Inter prediction in HEVC/H.265

Each inter-predicted PU has motion parameters for one or two reference picture lists. The motion parameters include a motion vector and a reference picture index. The use of one of the two reference picture lists may be signaled using inter_pred_idc. The motion vector may be explicitly coded as a differential related to the predictor, and such a coding mode is called AMVP mode.

When a CU is coded in skip mode, one PU is associated with this CU, there are no significant residual coefficients, no coded motion vector differentials, and no reference picture indexes. A merge mode is specified, whereby motion parameters for the current PU are obtained from neighboring PUs, including spatial and temporal candidates. The merge mode can be applied to any inter-predicted PU, not just for skip mode. An alternative to the merge mode is an explicit transmission of motion parameters, explicitly signaling per PU the motion vectors corresponding to each reference picture list and the reference picture index for the use of the reference picture list.

If the signaling indicates to use one of two reference picture lists, generate the PU from one block of samples. This is called "uni-prediction". Uni-prediction is available for both P and B slices.

If the signaling indicates to use both reference picture lists, generate a PU from two blocks of samples. This is called "bi-prediction". Bi-prediction is available for B slices only.

The inter prediction modes defined in HEVC are explained in detail below. First, merge mode is explained.

2.2.1 Merge mode

2.2.1.1 Deriving merge mode candidates

When predicting a PU using merge mode, we parse an index from the bitstream that points to an entry in the merge candidate list and use it to look up the motion information. The construction of this list is specified in the HEVC standard and can be summarized based on the following sequence of steps:
Step 1: Derive initial candidates o Step 1.1: Derive spatial candidates o Step 1.2: Redundancy check of spatial candidates o Step 1.3: Derive temporal candidates Step 2: Insert additional candidates o Step 2.1: Create bi-prediction candidates o Step 2.2: Insert zero motion candidates

These steps are also shown diagrammatically in Fig. 9. For spatial merge candidate derivation, select up to four merge candidates among candidates at five different positions. For temporal merge candidate derivation, select up to one merge candidate among two candidates. Since the decoder side assumes a constant number of candidates per PU, generate additional candidates if the number of candidates does not reach the maximum number of merge candidates (MaxNumMergeCand) signaled in the slice header. Since the number of candidates is constant, the index of the best merge candidate is coded using truncated unary binarization (TU). If the size of the CU is equal to 8, all PUs of the current CU share one merge candidate list, which is the same as the merge candidate list of the 2Nx2N prediction unit.

The operations associated with the steps above are explained in detail below.

2.2.1.2 Deriving spatial candidates

In the derivation of spatial merging candidates, up to four merging candidates are selected from the candidates at the positions shown in FIG. 10. The order of derivation is _A1 , _B1 , _B0 , _A0 , _B2 . Position B2 is considered only if any PU at positions _A1 , _B1 , _B0 , _A0 is unavailable (e.g., because it belongs to another slice or tile) or is intra _- coded. After adding the candidate at position _A1 , adding the remaining candidates is subjected to a redundancy check, which can ensure that candidates with the same motion information are removed from the list and improve coding efficiency. To reduce computational complexity, the aforementioned redundancy check does not consider all possible candidate pairs. Instead, it considers only the pairs linked by arrows in FIG. 11 and adds a candidate to the list only if the corresponding candidate used for the redundancy check does not have the same motion information. Another source of duplicated motion information is a "second PU" associated with a partition different from 2N×2N. As an example, Fig. 12 shows the second PU for Nx2N and 2NxN, respectively. When splitting the current PU into Nx2N, the candidate at position _A1 is not considered for list construction. In fact, adding this candidate would lead to two prediction units with the same motion information, which is redundant for having only one PU in one coding unit. Similarly, when splitting the current PU into 2NxN, position _B1 is not considered.

2.2.1.3 Deriving temporal candidates

In this step, only one candidate is added to the list. Specifically, in the derivation of this temporal merge candidate, a scaled motion vector is derived based on the co-located PU belonging to the picture with the smallest POC difference with the current picture in a given reference picture list. In the slice header, the reference picture list used for the derivation of the co-located PU is explicitly signaled. As shown by the dotted line in Figure 13, the scaled motion vector of the temporal merge candidate is obtained, which is scaled from the motion vector of the co-located PU using the POC distances tb and td. tb is defined as the POC difference between the reference picture of the current picture and the current picture, and td is defined as the POC difference between the reference picture of the co-located PU and the co-located picture. The reference picture index of the temporal merge candidate is set equal to zero. The practical realization of this scaling process is described in the HEVC specification. For B slices, we take two motion vectors, one for reference picture list 0 and one for reference picture list 1, and combine them to form a bi-predictive merge candidate. Description of motion vector scaling for temporal merge candidates.

For the co-located PU(Y) belonging to the reference frame, select a location of the temporal candidate between candidates _C0 and _C1 as shown in Fig. 14. If the PU at location _C0 is not available, is intra-coded, or is outside the current CTU, location _C1 is used. Otherwise, location _C0 is used to derive the temporal merging candidate.

2.2.1.4 Inserting additional candidates

Besides spatial-temporal merge candidates, there are two additional types of merge candidates: joint bi-predictive merge candidates and zero merge candidates. The joint bi-predictive merge candidates are utilized to generate joint bi-predictive merge candidates. The joint bi-predictive merge candidates are used only for B slices. The joint bi-predictive candidate is generated by combining the first reference picture list motion parameters of the first candidate with the second reference picture list motion parameters of another candidate. If these two tuples provide different motion hypotheses, they form a new bi-predictive candidate. As an example, FIG. 15 shows the case where two candidates with mvL0 and refIdxL0 or mvL1 and refIdxL1 in the original list (left) are used to generate a joint bi-predictive merge candidate that is added to the final list (right). There are various rules for the combinations considered to generate these additional merge candidates, as specified here.

By inserting zero motion candidates and filling the remaining entries in the merge candidate list, the MaxNumMergeCand capacity is hit. These candidates have a spatial displacement of zero and a reference picture index that starts from zero and increases each time a new zero motion candidate is added to the list. The number of reference frames used by these candidates is one for unidirectional prediction and two for bidirectional prediction, respectively. Finally, no redundancy check is performed on these candidates.

2.2.1.5 Motion estimation regions for parallel processing

To speed up the encoding process, motion estimation can be done in parallel, thereby deriving motion vectors for all prediction units in a given region simultaneously. Since one prediction unit cannot derive motion parameters from neighboring PUs until its associated motion estimation is completed, the derivation of merge candidates from spatial neighborhoods can interfere with parallel processing. To mitigate the tradeoff between coding efficiency and processing latency, HEVC specifies a Motion Estimation Region (MER), whose size is signaled in the picture parameter set using the "log2_parallel_merge_level_minus2" syntax element. When specifying one MER, merge candidates in the same region are marked as unavailable and therefore not considered in the list construction.
7.3.2.3 Picture parameter set RBSP syntax 7.3.2.3.1 General picture parameter set RBSP syntax

log2_parallel_merge_level_minus2+2 specifies the value of the variable Log2ParMrgLevel used in the derivation process of luminance motion vectors in the merge mode specified in Section 8.5.3.2.2.2 and in the derivation process of spatial merge candidates specified in Section 8.5.3.2.3. The value of log2_parallel_merge_level_minus2 is in the range of 0 to CtbLog2SizeY-2 inclusive.
The variable Log2ParMrgLevel is derived as follows:
Log2ParMrgLevel=log2_parallel_merge_level_minus2+2 (7-37)
NOTE 3: The value of Log2ParMrgLevel indicates the built-in capability of deriving merge candidate lists in parallel. For example, if Log2ParMrgLevel is equal to 6, then merge candidate lists for all prediction units (PUs) and coding units (CUs) contained in a 64x64 block can be derived in parallel.

2.2.2 Motion vector prediction in AMVP mode

Motion vector prediction exploits the spatial-temporal correlation between motion vectors and neighboring PUs, which is used for explicit transmission of motion parameters. First, a motion vector candidate list is constructed by checking the availability of the left, upper temporally neighboring PU positions, removing redundant candidates, and adding zero vectors to keep the length of the candidate list constant. Then, the encoder can select the best predictor from the candidate list and transmit the corresponding index indicating the selected candidate. Similar to the merge index signaling, the index of the best motion vector candidate is coded using a shortened unary term. The maximum value to be coded in this case is 2 (e.g., Figures 2 to 8). The following sections will explain the details of the motion vector prediction candidate derivation process.

2.2.2.1 Deriving motion vector prediction candidates

Figure 16 summarizes the process of deriving motion vector prediction candidates.

In motion vector prediction, two types of motion vector candidates are considered: spatial motion vector candidates and temporal motion vector candidates. To derive spatial motion vector candidates, two motion vector candidates are ultimately derived based on the motion vectors of each PU at five different positions, as shown in FIG. 11.

For derivation of a temporal motion vector candidate, select one motion vector candidate from two candidates derived based on two different co-location arrangements. After creating an initial list of spatial-temporal candidates, remove duplicate motion vector candidates in the list. If the number of possible candidates is greater than two, remove from the list the motion vector candidates whose reference picture index in the associated reference picture list is greater than 1. If the number of spatial-temporal motion vector candidates is less than two, add an additional zero motion vector candidate to the list.

2.2.2.2 Spatial motion vector candidates

In deriving spatial motion vector candidates, we consider up to two candidates out of five possible candidates derived from PUs located as shown in Fig. 11 that are in the same position as the motion merge. The order of derivation for the left side of the current PU is defined as _A0 , _A1 , scaled _A0 , scaled _A1 . The order of derivation for the top side of the current PU is defined as _B0 , _B1 , _B2 , scaled _B0 , scaled _B1 , scaled _B2 . So, for each side, there are four possible cases for motion vector candidates: two cases that do not require spatial scaling and two cases that use spatial scaling. The four different cases can be summarized as follows:
No spatial scaling: (1) Same reference picture list and same reference picture index (same POC)
- (2) Different reference picture lists and the same reference picture (same POC)
Spatial Scaling - (3) Same reference picture list but different reference pictures (different POC)
(4) Different reference picture lists and different reference pictures (different POCs)

Check the non-spatial scaling case first, then do spatial scaling. Consider spatial scaling if POC is different between neighboring PU's reference picture and current PU's reference picture, regardless of reference picture list. Scaling top motion vector helps parallel derivation of left and top MV candidates if all PUs of left candidate are not available or are intra-coded. Otherwise, no spatial scaling is allowed for top motion vector.

In the spatial scaling process, we scale the motion vectors of neighboring PUs in a similar manner to temporal scaling, as shown in Figure 17. The main difference is that the reference picture list and index of the current PU are given as input, and the actual scaling process is the same as temporal scaling.

2.2.2.3 Temporal motion vector candidates

Other than deriving the reference picture index, all the processes for deriving temporal merge candidates are the same as the processes for deriving spatial motion vector candidates (see Figure 6). The reference picture index is signaled to the decoder.

2.2.2.4 Signaling of AMVP information

For AMVP mode, four parts can be signaled in the bitstream: prediction direction, reference index, MVD, and mv predictor candidate index.
Syntax table:

7.3.8.9 Motion Vector Difference Syntax

2.3 New inter-prediction method in JEM (Joint Exploration Model)

2.3.1 Sub-CU based motion vector prediction

In JEM with QTBT, each CU can have at most one set of motion parameters for each prediction direction. In the encoder, we consider two sub-CU level motion vector prediction methods by splitting a large CU into sub-CUs and deriving motion information for all sub-CUs of the large CU. The Alternative Temporal Motion Vector Prediction (ATMVP) method allows each CU to fetch multiple sets of motion information from multiple blocks smaller than the current CU in the aligned reference picture. In the Spatial-Temporal Motion Vector Prediction (STMVP) method, we use the temporal motion vector predictor and spatial neighborhood motion vectors to recursively derive the motion vectors of the sub-CUs.

To maintain a more accurate motion field for sub-CU motion estimation, reference frame motion compression is currently disabled.

2.3.1.1 Alternative temporal motion vector prediction

In Alternative Temporal Motion Vector Prediction (ATMVP), the Temporal Motion Vector Prediction (TMVP) method is modified by fetching multiple sets of motion information (including motion vectors and reference indexes) from blocks smaller than the current CU. As shown in Figure 18, a sub-CU is a square NxN block (by default, N is set to 4).

ATMVP predicts motion vectors of sub-CUs in a CU in two steps. The first step is to identify the corresponding block in a reference picture with a so-called temporal vector. This reference picture is called the motion source picture. The second step is to split the current CU into sub-CUs and obtain the motion vector and reference index of each sub-CU from the block corresponding to each sub-CU, as shown in Figure 18.

In the first step, the reference picture and corresponding block are determined by the motion information of the spatially neighboring blocks of the current CU. To avoid the repeated scanning process of the neighboring blocks, the first merging candidate in the merging candidate list of the current CU is used. The first available motion vector and its associated reference index are set to the temporal vector and the index of the motion source picture. In this way, the corresponding block can be identified more accurately in ATMVP compared with TMVP, and the corresponding block (sometimes called the aligned block) is always in the lower right or center position with respect to the current CU. In one example, if the first merging candidate is from the left neighboring block (i.e., A ₁ in FIG. 19 ), the associated MV and reference picture are utilized to identify the source block and source picture.

Figure 19 shows specific examples of source blocks and source pictures.

In the second step, the corresponding block of the sub-CU is identified by its temporal vector in the motion source picture by adding the temporal vector to the coordinates of the current CU. For each sub-CU, the motion information of its corresponding block (the smallest motion grid covering the center sample) is used to derive the motion information of the sub-CU. After identifying the motion information of the corresponding N×N block, it is converted into the motion vector and reference index of the current sub-CU, and motion scaling and other procedures are applied, similar to TMVP in HEVC. For example, the decoder checks whether the low-latency condition (i.e., the POC of all reference pictures of the current picture is smaller than the POC of the current picture) is met, and possibly predicts the motion vector MVy (X equals 0 or 1, and Y equals 1-X) of each sub-CU using the motion vector MVx (the motion vector corresponding to the reference picture list X).

2.3.1.2 Spatial-temporal motion vector prediction

In this method, motion vectors for sub-CUs are derived recursively along the raster scan order. Figure 20 illustrates this concept. Consider an 8x8 CU that contains four 4x4 sub-CUs, A, B, C, and D. The neighboring 4x4 blocks in the current frame are labeled a, b, c, and d.

The motion derivation of sub-CU A starts by identifying its two spatial neighbors. The first neighbor is the N×N block above sub-CU A (block c). If this block c is not available or is intra-coded, check the other N×N blocks above sub-CU A (starting from block c, going from left to right). The second neighbor is the block to the left of sub-CU A (block b). If block b is not available or is intra-coded, check the other blocks to the left of sub-CU A (starting from block b, going from top to bottom). The motion information obtained from the neighboring blocks in each list is scaled to the first reference frame of a given list. Then, derive the TMVP for sub-block A following a procedure similar to the TMVP derivation specified in HEVC. Fetch the motion information of the aligned blocks at position D and scale accordingly. Finally, after retrieving and scaling the motion information, we average all available motion vectors (up to 3) separately for each reference list. Let this averaged motion vector be the motion vector of the current sub-CU.

Figure 20 shows an example of one CU with four subblocks (A-D) and their neighboring blocks.

2.3.1.3 Sub-CU motion prediction mode signal notification

Sub-CU mode is enabled as an additional merge candidate, and no additional syntax elements are required to signal the mode. Add two additional merge candidates to the merge candidate list of each CU to represent ATMVP and STMVP modes. Use up to seven merge candidates if the sequence parameter set indicates that ATMVP and STMVP are enabled. The encoding logic of the additional merge candidates is the same as that of the merge candidates in HM, i.e., for each CU in a P or B slice, two or more RD checks are required for the two additional merge candidates.

In JEM, all bins of the merge index are CABAC coded contexts, whereas in HEVC, only the first bin is a coded context and the remaining bins are bypass coded contexts.

2.3.2 Adaptive motion vector difference resolution

In HEVC, when use_integer_mv_flag is 0 in the slice header, the Motion Vector Difference (MVD) (the difference between the motion vector and the predicted motion vector of the PU) is signaled in units of 1/4 luma samples. In JEM, Locally Adaptive Motion Vector Resolution (LAMVR) is introduced. In JEM, MVD can be decoded in units of 1/4 luma samples, integer luma samples, or 4 luma samples. MVD resolution is controlled at the coding unit (CU) level, and the MVD resolution flag is conditionally signaled for each CU that has at least one non-zero MVD component.

For a CU with at least one non-zero MVD component, a first flag is signaled to indicate whether quarter luma sample MV precision is used in the CU. If the first flag (equal to 1) indicates that quarter luma sample MV precision is not used, another flag is signaled to indicate whether integer luma sample MV precision or 4 luma sample MV precision is used.

If the first MVD resolution flag of a CU is zero or is not coded for the CU (i.e., all MVDs in the CU are zero), then 1/4 luma sample MV resolution is used for the CU. If the CU uses integer luma sample MV precision or 4 luma sample MV precision, round the MVPs in the CU's AMVP candidate list to the corresponding precision.

In the encoder, the CU-level RD check is used to determine which MVD resolution to use for the CU. That is, the CU-level RD check is performed three times for each MVD resolution. To increase the encoder speed, the following encoding method is applied in JEM:

During RD check of a CU with normal 1/4 luma sample MVD resolution, the motion information (integer luma sample precision) of the current CU is stored. During RD check of the same CU with integer luma sample and 4 luma sample MVD resolution, the stored motion information (after rounding) is used as a starting point for further small range motion vector refinement, so that the time consuming motion estimation process is not duplicated three times.

Conditionally invoke RD check for CUs with 4 luma sample MVD resolution. For a CU, if the RD cost of integer luma sample MVD resolution is much greater than that of 1/4 luma sample MVD resolution, then the RD check of 4 luma sample MVD resolution for the CU is omitted.

2.3.3 Pattern matching motion vector derivation

PMMVD (Pattern Matched Motion Vector Derivation) mode is a special merge mode based on FRUC (Frame-Rate Up Conversion) technology. In this mode, the motion information of the blocks is not signaled but derived on the decoder side.

If that merge flag is true, then the FRUC flag is signaled to the CU. If the FRUC flag is false, then the merge index is signaled and normal merge mode is used. If the FRUC flag is true, then an additional FRUC mode flag is signaled to indicate which method (bilateral matching or template matching) is used to derive the motion information for the block.

On the encoder side, the decision to use FRUC merge mode for a CU is based on the RD cost selection, just as it is done for normal merge candidates. That is, we check both two matching modes (bilateral matching and template matching) for one CU using the RD cost selection. The one that leads to the minimum cost is further compared with other CU modes. If the FRUC matching mode is the most efficient one, we set the FRUC flag to true for the CU and use the associated matching mode.

The motion derivation process in FRUC merge mode has two steps. First, a CU-level motion search is performed, and then a sub-CU-level motion refinement is performed. At the CU level, an initial motion vector for the entire CU is derived based on bilateral matching or template matching. First, a list of MV candidates is generated, and the candidate that leads to the minimum matching cost is selected as the opening starting point for further CU-level refinement. Then, a local search based on bilateral matching or template matching is performed near the starting point, and the MV result that results in the minimum matching cost is taken as the MV for the entire CU. Next, the derived CU motion vector is used as the starting point to further refine the motion information at the sub-CU level.

For example, to derive WxH CU motion information, the following derivation process is performed: In the first stage, MVs for the entire WxH CU are derived. In the second stage, the CU is further divided into MxM sub-CUs. The value of M is calculated as in (16), where D is a predefined division depth, which is set to 3 by default in JEM. Then, MVs for each sub-CU are derived.

As shown in Figure 21, this bilateral matching is used to derive the motion information of the current CU by finding the closest match between two blocks along the motion trajectory of the current CU in two different reference pictures. Assuming a continuous motion trajectory, the motion vectors MV0 and MV1 pointing to the two reference blocks are proportional to the temporal distance between the current picture and the two reference pictures, e.g., TD0 and TD1. As a special case, when the current picture is temporally between the two reference pictures and the temporal distance from the current picture to the two reference pictures is the same, the bilateral matching becomes a mirror-based bidirectional MV.

As shown in FIG. 22, template matching is used to derive motion information of the current CU by finding the closest match between a template in the current picture (nearby blocks above and/or to the left of the current CU) and a block in the reference picture (same size as the template). Besides the aforementioned FRUC merge mode, template matching is also applied to the AMVP mode. In JEM, similar to HEVC, AMVP has two candidates. A new candidate is derived by using the template matching method. If the newly derived candidate by template matching is different from the first existing AMVP candidate, it is inserted at the beginning of the AMVP candidate list, and then the list size is set to 2 (meaning removing the second existing AMVP candidate). When applied to the AMVP mode, only CU level search is applied.

2.3.3.1 CU-level MV candidate set

The CU-level MV candidate set consists of:
(i) the original AMVP candidate if the current CU is in AMVP mode; (ii) all merge candidates;
(iii) Multiple MVs in an interpolated MV field.
(iv) Motion vectors of the top and left neighbors

When bilateral matching is used, each valid MV of the merge candidate is used as input to generate an MV pair assuming bilateral matching. For example, one valid MV of the merge candidate is (MVa, refa) in reference list A. Then, the reference picture refb of the paired bilateral MV is found in another reference list B, where refa and refb are on different sides of the current picture in time. If no such refb is available in reference list B, determine refb as a reference different from refa, whose temporal distance to the current picture is the minimum value in list B. After determining refb, derive MVb by scaling MVa based on the temporal distance between the current picture and refa and refb.

Add the four MVs from the interpolated MV field to the CU level candidate list as well. Specifically, add the interpolated MVs at positions (0,0), (W/2,0), (0,H/2), and (W/2,H/2) of the current CU.

When applying FRUC in AMVP mode, the original AMVP candidate is also added to the CU-level MV candidate set.

At the CU level, add up to 15 MVs for AMVP CU and up to 13 MVs for merged CU to the candidate list.

2.3.3.2 Sub-CU level MV candidate set

The sub-CU level MV candidate set consists of:
(i) MVs determined from CU-level searches;
(ii) MVs in the top, left, top left, and top right neighborhoods;
(iii) a scaled version of the collocated MV from the reference picture;
(iv) up to four ATMVP candidates;
(v) Up to four STMVP candidates

The scaled MV from the reference picture is derived as follows: traverse all reference pictures in both lists. The MV at the sub-CU's alignment position in the reference picture is scaled with respect to the reference of the starting CU level MV.

ATMVP and STMVP candidates are limited to the first four candidates

At the sub-CU level, a maximum of 17 MVs are added to the candidate list.

2.3.3.3 Generation of interpolated MV fields

Before encoding a frame, an interpolated motion field is generated for the entire picture based on one ME. This motion field may then be used as a CU-level or sub-CU-level MV candidate.

First, the motion field of each reference picture in both reference lists is traversed at the 4x4 block level. For each 4x4 block, if the motion associated with the block passing through the 4x4 block in the current picture (shown in Figure 23) does not already have an interpolated motion assigned to it, scale the motion of the reference block to the current picture based on the temporal distances TD0 and TD1 (similar to MV scaling of TMVP in HEVC) and assign the scaled motion to the block in the current frame. If the 4x4 block does not have a scaled MV assigned to it, the motion of the block is marked as unavailable in the interpolated motion field.

2.3.3.4 Interpolation and matching costs

When one motion vector points to one fractional sample location, motion compensated interpolation is needed. To reduce complexity, we use bilinear interpolation for both bilateral and template matching instead of the usual 8-tap HEVC interpolation.

The calculation of the matching cost is slightly different in different steps. When selecting a candidate from the CU-level candidate set, the matching cost is the sum of absolute differences (SAD) of bilateral matching or template matching. After determining the starting MV, we calculate the matching cost C of bilateral matching in the sub-CU level search as follows:

where w is a weighting factor empirically set to 4, and MV and MV ^s denote the current MV and the starting MV, respectively. The SAD is still used as the matching cost for template matching in the sub-CU level search.

In FRUC mode, MV is derived by using only luma samples. The derived motion is used for both luma and chroma for MC inter prediction. After determining MV, the final MC is performed using an 8-tap interpolation filter for luma and a 4-tap interpolation filter for chroma.

2.3.3.5 MV improvements

MV refinement is a pattern-based MV search with the criteria of bilateral matching cost or template matching cost. JEM supports two search patterns, namely Unrestricted Center-Biased Diamond Search (UCBDS) and adaptive cross-search for MV refinement at CU level and sub-CU level, respectively. For both CU and sub-CU level MV refinement, MVs are directly searched with the accuracy of 1/4 luma sample MV, followed by refinement of 1/8 luma sample MV. The search range of MV refinement for CU and sub-CU steps is set equal to 8 luma samples.

2.3.3.6 Prediction direction selection in template matching FRUC merge mode

In the bilateral matching merge mode, bidirectional prediction is always applied to derive the motion information of a CU based on the closest matching between two blocks along the motion trajectory of the current CU in two different reference pictures. For the template matching merge mode, there is no such limitation. In the template matching merge mode, the encoder can select among uni-prediction from list0, uni-prediction from list1, or bi-prediction for a CU. The selection is made based on the template matching cost as follows:
If costBi≦factor*min(cost0, cost1), then use bidirectional prediction.
Otherwise, if cost0≦cost1, use single prediction from list0.
If not,
Use a single prediction from list1.

Here, cost0 is the SAD of list0 template matching, cost1 is the SAD of list1 template matching, and costBi is the SAD of bidirectional prediction template matching. When the factor value is 1.25, it means that the selection process is biased towards bidirectional prediction.
This inter prediction direction selection is applied only to the template matching process at the CU level.

2.3.4 Decoder-side motion vector improvements

In bidirectional prediction operations, bi-predictive blocks constructed using the motion vectors (MVs) in list0 and list1, respectively, are combined to form one prediction signal to predict one block region. In the Decoder-side Motion Vector Refinement (DMVR) method, the two motion vectors of bidirectional prediction are further refined by a bilateral template matching process. To obtain the refined MVs without transmitting additional motion information, bilateral template matching is applied in the decoder to perform a distortion-based search between the bilateral template and the reconstructed samples in the reference picture.

In DMVR, we generate a bilateral template from the first MV0 in list0 and MV1 in list1 as a weighted combination (i.e., average) of two prediction blocks, respectively, as shown in Fig. 23. The template matching operation consists of calculating a cost measure between the generated template and a sample region in the reference picture (near the first prediction block). For each of the two reference pictures, the MV with the smallest template cost is considered as the updated MV of that list and replaces the original MV. In JEM, we search for nine MV candidates for each list. The nine MV candidates include the original MV and eight surrounding MVs with one luma sample offset with respect to the original MV in either horizontal or vertical directions or both directions. Finally, we generate the final bidirectional prediction result using two new MVs, i.e., MV0' and MV1' as shown in Fig. 24. We use the sum of absolute differences (SAD) as the cost measure.

DMVR applies to bi-prediction merge mode between one MV from a past reference picture and one MV from a future reference picture without transmitting additional syntax elements. In JEM, DMVR is not applied if LIC, affine motion, FRUC, or sub-CU merge candidates are enabled for the CU.

2.3.5 Merge/skip mode with improved bilateral matching

First, a merge candidate list is constructed by inserting motion vectors and reference indices of spatially and temporally neighboring blocks into a candidate list with redundancy check until the number of available candidates reaches the maximum candidate size of 19. The merge candidate list for merge/skip mode is constructed by inserting spatial candidates (FIG. 11) used for HEVC (merge candidates and zero candidates), temporal candidates, affine candidates, Advanced Temporal MVP (ATMVP) candidates, Spatial Temporal MVP (STMVP) candidates, and additional candidates based on a predefined insertion order.

- Spatial candidates for blocks 1 to 4

- Extrapolation affine candidates for blocks 1 to 4

-ATMVP

-STMVP

- Virtual affine candidates

- Spatial Candidates (Block 5) (used only if the number of available candidates is less than 6).

- Extrapolation affine candidates (Block 5)

-Temporal candidates (derived like HEVC)

- Non-adjacent spatial candidates following the extrapolated affine candidates (blocks 6 to 49 shown in Figure 25).

- Candidates for joining

-Zero candidates

Note that the IC flag is inherited from the merge candidate, except for STMVP and affine. Also, for the first four spatial candidates, the bi-predictive candidates are inserted before the uni-predictive ones.

In some embodiments, blocks that are not connected to the current block can be accessed. If the non-adjacent blocks are coded in a non-intra mode, the associated motion information may be added as additional merging candidates.

2.3.6 Shared merge list JVET-M0170

To enable parallel processing of small skip/merge coded CUs, we propose to share the same merge candidate list for all leaf coding units (CUs) of an ancestor node in the CU partition tree. We call the ancestor node a merge-share node. We generate a shared merge candidate list at the merge-share node, pretending that the merge-share node is a leaf CU.

In the definition of Type-2, during the parsing stage of decoding, a merge-shared node is determined for each CU within a CTU. In addition, the merge-shared node is an ancestor node of the leaf CU, and must satisfy the following two criteria:

The size of the merge shared node must be greater than or equal to the size threshold.

In a merge shared node, the size of the child CU is smaller than the size threshold.

Furthermore, we need to ensure that the samples of a merge-shared node are not outside the picture boundary. During the parsing phase, if an ancestor node satisfies criteria (1) and (2) but has some samples outside the picture boundary, then this ancestor node is not a merge-shared node, and we go ahead and find a merge-shared node for its child CU.

Figure 35 shows an example of the difference between Type-1 and Type-2 definitions. In this example, the parent node is split into three child CUs. The size of the parent node is 128. In the Type-1 definition, the three child CUs are separate merge-shared nodes. However, in the Type-2 definition, the parent node is a merge-shared node.

The proposed shared merge candidate list algorithm supports translational merge (including merge mode and triangle merge mode, history-based candidates are also supported) and subblock-based merge modes. In all kinds of merge modes, the behavior of the shared merge candidate list algorithm basically looks the same, generating candidates in the merge share node that only make the merge share node look like a leaf CU. It has two major advantages. The first advantage is that it enables parallel processing for merge modes, and the second advantage is that it shares all the computations of all leaf CUs in the merge share node. Therefore, it can greatly reduce the hardware cost of all merge modes for hardware codecs. The proposed shared merge candidate list algorithm allows the encoder and decoder to easily accommodate parallel encoding of merge modes, and alleviates the cycle budget problem of merge modes.

2.3.7 Tile groups

In JVET-L0686, slices are removed in favor of tile groups, and the HEVC syntax element slice_address is replaced with the tile_group_address in the tile_group_header as the address of the first tile in the tile group (if there are multiple tiles in the picture).

3. Examples of problems that the embodiments disclosed herein aim to solve

Current HEVC design can correlate the neighboring blocks of the current block (next to the current block) to better code the motion information. However, the neighboring blocks may correspond to different objects with different motion trajectories. In this case, prediction from the neighboring blocks is not efficient.

Prediction from motion information of non-adjacent blocks incurs the cost of storing all the motion information (typically at 4x4 level) in a cache, which provides no additional coding gain and significantly increases the complexity of the hardware implementation.

4. Some examples

To overcome the shortcomings of existing implementations, various embodiments may implement a LUT-based motion vector prediction technique that uses one or more tables (e.g., lookup tables) in which at least one motion candidate is stored to predict motion information of a block, providing video coding with higher coding efficiency. The lookup table is one example of a table that may be used to include motion candidates to predict motion information of a block, and other implementations are possible. Each LUT may include one or more motion candidates, each associated with corresponding motion information. The motion information of a motion candidate may include some or all of a prediction direction, a reference index/picture, a motion vector, a LIC flag, an affine flag, a Motion Vector Derivation (MVD) accuracy, and/or a MVD value. The motion information may further include block position information to indicate where the motion information originates.

The LUT-based motion vector prediction based on the disclosed technology can improve both existing and future video coding standards, and is elucidated in the following examples for various embodiments. Since the LUT enables the encoding/decoding process to be performed based on history data (e.g., already processed blocks), the LUT-based motion vector prediction can also be called HMVP History-based Motion Vector Prediction (HMVP) method. In the LUT-based motion vector prediction method, one or more tables with motion information from previously coded blocks are maintained during the encoding/decoding process. These motion candidates stored in the LUT are called HMVP candidates. During the encoding/decoding of a block, the associated motion information in the LUT may be added to a motion candidate list (e.g., merge/AMVP candidate list) to use the LUT after encoding/decoding a block. The updated LUT is then used to code the subsequent block. In this way, the update of the motion candidates in the LUT is based on the encoding/decoding order of the blocks. The following examples should be considered as examples to explain the general concept. These examples should not be interpreted in a narrow sense. Moreover, these examples can be combined in any way.

In some embodiments, one or more lookup tables in which at least one motion candidate is stored may be used to predict motion information for a block. An embodiment may use a motion candidate to indicate a set of motion information stored in the lookup table. In the case of a conventional AMVP or merge mode, an embodiment may use an AMVP or merge candidate to store the motion information.

The following examples illustrate the general concept.

Lookup table example

Example A1: Each lookup table may contain one or more motion candidates, with each candidate associated with its motion information.
The motion information of a motion candidate here may include some or all of the following: prediction direction, reference index/picture, motion vector, LIC flag, affine flag, MVD precision, MVD value.
b. The motion information may further include block position information and/or block shape to indicate where the motion information comes from.

Selecting a LUT

Example B1: When encoding a block, some or all of the motion candidates from a lookup table may be checked in sequence. When a motion candidate is checked while encoding a block, this motion candidate may be added to a motion candidate list (e.g., AMVP, merge candidate list). Example B2: The selection of the lookup table may depend on the position of the block.

How to use a lookup table

Example C1: The total number of motion candidates in the lookup table to be checked may be predefined.

Example C2: One or more motion candidates contained in one lookup table may be directly inherited by one block.
They may be used for merge mode coding, i.e., motion candidates may be checked in the merge candidate list derivation process.
b. They may be used for affine merge mode encoding.
i. If the affine flag is 1, then the motion candidates in the lookup table can be added as affine merge candidates.
c) They may be used for other kinds of merge modes such as sub-block merge mode, affine merge mode, triangle merge mode, inter-intra merge mode, MMVD (Merge with MVD) mode.
d. Checking of motion candidates in the lookup table may be enabled if:
i. After inserting the TMVP candidate, the merge candidate list is not full.
ii. After checking certain spatially neighboring blocks for spatial merge candidate derivation, the merge candidate list is not full.
iii. After all spatial merge candidates, the merge candidate list is not full.
iv. After the combined bi-predictive merge candidates, the merge candidate list is not full.
v. If the number of spatial or temporal (e.g., including spatially adjacent and non-spatially adjacent, TMVP, STMVP, ATMVP, etc.) merge candidates put into the merge candidate list from other coding schemes (e.g., the merge derivation process of the HEVC design, or the JEM design) is less than the maximum allowed merge candidates minus a given threshold.
1. In one example, the threshold is set to 1 or 0.
2. Alternatively, the threshold may be signaled in the SPS/PPS/sequence, picture, slice header/tile or may be pre-defined.
3. Alternatively, the threshold may be adaptively changed per block, e.g., it may depend on coded block information like block size/block shape/slice type and/or it may depend on the number of available spatial or temporal merging candidates.
4. In another example, the number of certain merge candidates that are not already included in the merge candidate list is less than the maximum allowed merge candidates minus a given threshold. The "certain merge candidates" may be spatial candidates like HEVC or non-adjacent merge candidates.
vi. Pruning may be applied before adding motion candidates to the merge candidate list. In various implementations of this and other examples disclosed in this patent specification, pruning may include a) comparing the motion information with existing entries for uniqueness, or b) adding the motion information to the list if unique, or c) if not unique, c1) not adding the motion information, or c2) adding the motion information and removing the matching existing entry. In some implementations, no pruning step is performed when adding motion candidates from the table to the candidate list.
1. In one example, motion candidates may be pruned to all or a portion of spatial or temporal (e.g., including adjacent and non-adjacent spatial, TMVP, STMVP, ATMVP, etc.) merge candidates available from other coding methods in the merge candidate list.
2. Motion candidates may not be pruned to sub-block based motion candidates, e.g., ATMVP, STMVP.
3. In one example, the current motion candidate may be pruned to all or a portion of the available motion candidates in the merge candidate list (inserted before the current motion candidate).
4. The number of pruning steps associated with a motion candidate (e.g., the number of times a motion candidate needs to be compared with other candidates in the merge list) may depend on the number of available spatial or temporal merge candidates. For example, when checking a new motion candidate, if there are M candidates available in the merge list, the new motion candidate can only be compared with the first K candidates (K≦M). If the pruning function returns false (e.g., not identical to any of the first K candidates), the new motion candidate is considered to be different from all M candidates and can be added to the merge candidate list. In one example, K is set to min(K, 2).
5. In one example, only compare the newly added motion candidate with the first N candidates in the merge candidate list, for example N=3, 4 or 5. N may be signaled from the encoder to the decoder.
6. In one example, the new motion candidate being checked is only compared to the last N candidates in the merge candidate list, for example N=3, 4 or 5. N may be signaled from the encoder to the decoder.
7. In one example, the way in which candidates previously added to the list are selected from the table and compared to the new motion candidates may depend on where the previously added candidates were derived from.
In one example, motion candidates in a look-up table may be compared to candidates derived from given temporal and/or spatial neighboring blocks.
b. In one example, different entries of motion candidates in the lookup table may be compared to different candidates that were previously added (i.e., derived from different positions).
.
e. Checking of motion candidates in a lookup table may be enabled before checking other merge (or affine merge or other inter-coding method) candidates, such as those derived from adjacent/non-adjacent spatial or temporal blocks.
f. If there is at least one motion candidate in the lookup table, checking of motion candidates in the lookup table may be enabled.

Example C3: The motion candidates contained in the lookup table may be used as predictors for encoding the motion information of the block.
They may be used for AMVP mode coding, i.e., motion candidates may be checked in the AMVP candidate list derivation process.
b. They may be used for Symmetric Motion Vector Difference (SMVD) coding, which codes only a part of the MVD (e.g. only the MVD signaled for one reference picture list and derived from another reference picture list).
c. They may be used for Symmetric Motion Vector (SMV) coding, which codes only a part of the MVs (e.g. only those signaled for one reference picture list and derived from another reference picture list).
d. Checking of motion candidates in the lookup table may be enabled if:
i. After checking or inserting a TMVP candidate, the AMVP candidate list is not full.
ii. The AMVP candidate list is not full after selecting and pruning from spatial neighborhoods and immediately before inserting the TMVP candidate.
iii. If no AMVP candidates from the upper neighboring block exist without scaling, and/or no AMVP candidates from the left neighboring block exist without scaling.
iv. The AMVP candidate list is not full after inserting a particular AMVP candidate.
v. Before adding motion candidates to the AMVP candidate list, pruning may be applied.
vi. Rules similar to those described in vi. 3 and 4 of embodiment C2 may be applied to the AMVP mode.
e. Checking of motion candidates may be enabled before checking other AMVP (or SMVD/SMV/Affine Inter or other Inter coding methods) candidates, such as those derived from adjacent/non-adjacent spatial or temporal blocks.
f. If there is at least one motion candidate in the lookup table, motion candidate checking may be enabled.
g. Check for motion candidates that have the same reference picture as the current reference picture (i.e., have the same Picture-Order-Count (POC)). That is, if a motion candidate contains the same reference picture as the current reference picture, the corresponding motion vector may be considered for the AMVP candidate list construction process.
Alternatively, also check (in scaled MV) for motion candidates that have a different reference picture than the current reference picture, i.e. if a motion candidate has a different reference picture than the current reference picture, the corresponding motion vector may be considered for the AMVP candidate list building process.
Alternatively, first check all motion candidates that have the same reference picture as the current reference picture, and then check motion candidates that have a different reference picture from the current reference picture, i.e. assign a higher priority to motion candidates that have the same reference picture.
iii. Alternatively, motion candidates are checked in the merge as well.
iv. If one motion candidate is a bi-directional prediction candidate, it may first check the reference picture in reference picture list X (e.g., the reference picture index or picture order counter of the reference picture), and then check the reference picture in reference picture list Y (Y!=X, e.g., Y=1-X) if the current referenced picture list is X.
v. Alternatively, if one motion candidate is a bi-prediction candidate, one may first check the reference pictures (e.g., the reference picture index or picture order counter of the reference picture) in reference picture list Y (Y!=X, e.g., Y=1-X), and then check the reference pictures in reference picture list X if the current referenced picture list is X.
Alternatively, a reference picture in the reference picture list may be checked to see if it is an X associated with all the motion candidates to be checked before the reference picture in the reference picture list is an Y (Y!=X, e.g., Y=1-X) associated with all the motion candidates to be checked.

Example C4: The checking order of motion candidates in the lookup table is defined as follows (assuming that K (K≧1) motion candidates can be checked):
The last K motion candidates in the lookup table.
b. The first K % L candidates, where L is the size of the lookup table, where K>L.
c. If K≧L, then all candidates in the lookup table (L candidates).
d. Alternatively, it may also be based on descending order of motion candidate index.
e. Alternatively, it may also be based on the ascending order of the motion candidate index.
f) Alternatively, select the K motion candidates based on candidate information such as the distance of the current block to the position associated with the motion candidate.
i. In one example, select the K closest motion candidates.
ii. In one example, the candidate information may further take into account the shape of the block when calculating the distance.
g. In one example, the checking order of K motion candidates from a table containing L candidates may be defined as follows: _a0 , _a0 + _T0 , _a0 + _T0 + _T1 , a0+ _T0 + _T1 + _T2 , ... _a0 + _T0 + _T1 + _{T2 +} ...+ _TK-1 , selecting candidates with indices _a0 and T _i (i=0...K-1) in order, where a0 and T i are integer values.
i. In one example, _a0 is set to 0 (i.e., the first entry of the motion candidates in the table). Alternatively, _a0 is set to (K-L/K). The operation "/" is defined as integer division that truncates the result towards zero. Alternatively, _a0 is set to any integer between 0 and L/K.
1. Alternatively, the value of _a0 may depend on the coding information of the current block and neighboring blocks.
In one example, all the intervals T _i (i=0...K-1) are the same, say L/K. The operation "/" is defined as integer division that truncates the result towards zero.
In one example, (K, L, _a0 , _Ti ) is set to (4, 16, 0, 4), or (4, 12, 0, 3), or (4, 8, 0, 1), or (4, 16, 3, 4), or (4, 12, 2, 3), or (4, 8, 1, 2), where _Ti is the same for all i.
iv. Such methods may only be applied if K is less than L.
v. Alternatively, if K is greater than or equal to a threshold, part c of example C4 may be applied. The threshold may be defined as L, or may depend on K, or may be adaptively changed on a block-by-block basis. In one example, the threshold may depend on the number of available motion candidates in the list before adding new motion candidates from the lookup table.
h. In one example, the checking order of K motion candidates from a table containing L candidates may be defined as follows: _a0 , _a0 - _T0 , _a0 - _T0 - _T1 , _a0 - _T0 - _T1 - _T2 , ... _a0 - _T0 - _T1 - _T2 -...- _TK-1 , selecting candidates with indices _a0 and T _i (i is 0...K-1) in order.
In one example, a ₀ is set to L−1 (i.e., the last entry of the motion candidates in the table). Alternatively, a ₀ is set to any integer between L−1-L/K and L−1.
ii. In one example, all the intervals T _i (i=0...K-1) are the same, say L/K.
In one example, (K, L, a ₀ , T _i ) is set to (4, 16, L-1, 4), or (4, 12, L-1, 3), or (4, 8, L-1, 1), or (4, 16, L-4, 4), or (4, 12, L-3, 3), or (4, 8, L-2, 2), where _{T i} is the same for all i.
iv. Such methods may only be applied if K is less than L.
v. Alternatively, part c of example C4 may also be applied if K is greater than or equal to a threshold. The threshold may be defined as L, or may depend on K, or may be adaptively changed on a block-by-block basis. In one example, the threshold may depend on the number of available motion candidates in the list before adding new motion candidates from the lookup table.
i. The number and/or manner of selecting motion candidates from the look-table may depend on the coding information, e.g., block size/block shape.
i. In one example, for smaller block sizes, instead of selecting the last K motion candidates, one may select the other K motion candidates (not starting from the end).
ii. In one example, the coding information may be in AMVP mode or in merge mode.
iii. In one example, the coding information may be in affine mode or non-affine AMVP mode or non-affine merge mode.
iv. In one example, the coding information may be in affine AMVP (inter) mode, or affine merge mode, or non-affine AMVP mode, or non-affine merge mode.
v. In one example, the encoded information may or may not be in Current Picture Reference (CPR) mode.
vi. Alternatively, the method of selecting a motion candidate from the lookup table may further depend on the number of motion candidates in the lookup table and/or the number of available motion candidates in the list before adding a new motion candidate from the lookup table.
j. In one example, the maximum number of available motion candidates in the lookup table to be checked (i.e., that may be added to the merge/AMVP candidate list) may depend on the number of available motion candidates in the lookup table (indicated by N _avaiMCinLUT ) and/or the maximum number of allowed motion candidates to be added (which may be predefined or signaled) (indicated by NUM _maxMC ) and/or the number of available motion candidates in the candidate list before checking candidates from the lookup table (indicated by N _avaiC ).
i. In one example, the maximum number of motion candidates in the lookup table to be checked is set to the minimum of (N _avaiMCinLUT , NUM _maxMC , N _avaiC ).
ii. Alternatively, the maximum number of motion candidates in the lookup table to be checked is set to the minimum of (N _avaiMCinLUT , NUM _maxMC −N _avaiC ).
In one example, N _avaiC indicates the number of insertion candidates derived from spatially or temporally (adjacent and/or non-adjacent) neighboring blocks. Alternatively, the number of sub-block candidates (such as AMTVP, STMVP, etc.) is not counted in N _avaiC .
iv. _{NUM_maxMC} may depend on the coding mode, e.g., in merge mode and AMVP mode, _{NUM_maxMC} may be set to different values. In one example, for merge mode, _{NUM_maxMC} may be set to 4, 6, 8, 10, etc., and for AMVP mode, _{NUM_maxMC} may be set to 1, 2, 4, etc.
v. Alternatively, _{NUM_maxMC} may depend on other coding information such as block size, block shape, slice type, etc.
k. The order in which the different lookup tables are checked is specified in the next subsection, Lookup Table Usage.
l. Once the merge/AMVP candidate list reaches the maximum allowed number of candidates, this checking process ends.
m. Once the merge/AMVP candidate list reaches the maximum allowed number of candidates minus a threshold (Th), the checking process ends. In one example, Th may be predefined as a positive integer value, e.g., 1, 2, or 3. Alternatively, Th may be adaptively changed per block. Alternatively, Th may be signaled in the SPS/PPS/slice header, etc. Alternatively, Th may further depend on the block shape/block size/coding mode, etc. Alternatively, Th may depend on the number of available candidates before adding the motion candidates from the LUT.
Alternatively, it terminates when the number of added motion candidates reaches the maximum allowed number of motion candidates. The maximum allowed number of motion candidates may be signaled or predefined. Alternatively, the maximum allowed number of motion candidates may further depend on block shape/block size/coding mode etc.
o. One syntax element indicating the table size as well as the number of motion candidates that can be checked (i.e., K=L) may be signaled in the SPS, PPS, slice header, tile header.

In some implementations, the motion candidates in the lookup table may be used to derive other candidates, and the derived candidates may be used to encode the block.

In some implementations, the enable/disable of the use of a lookup table for motion information coding of a block may be signaled in the SPS, PPS, slice header, tile header, CTU, CTB, CU, or PU, an area covering multiple CTUs/CTBs/CUs/PUs.

In some implementations, whether to apply a prediction from the lookup table may further depend on the coding information. If it is inferred not to apply to a block, the additional signaling of the prediction indication is skipped. Alternatively, if it is inferred not to apply to a block, there is no need to access the motion candidates in the lookup table and the check of the associated motion candidates is omitted.

In some implementations, motion candidates in a look-up table in a previously coded frame/slice/tile may be used to predict motion information of a block in a different frame/slice/tile.
In one example, only the look-up table associated with the reference picture of the current block may be utilized to code the current block.
b. In one example, only look-up tables associated with pictures having the same slice type and/or the same quantization parameter of the current block may be utilized to code the current block.

Update lookup table

After encoding a block with motion information (i.e., IntraBC mode, inter coding mode), one or more lookup tables may be updated.

In all of the above examples and implementations, the lookup table indicates coded information, or information derived from coded information from previously coded blocks, in decoding order.
The lookup table may contain translational motion information, or affine motion information, or affine model parameters, or intra-mode information, or lighting compensation information, etc.
b. Alternatively, the lookup table may contain at least two types of information, such as translational motion information, or affine motion information, or affine model parameters, or intra-mode information, or lighting compensation information, etc.

Additional exemplary embodiments

A History-based MVP (HMVP) method is proposed, which defines HMVP candidates as motion information of previously coded blocks. During the coding/decoding process, a table with multiple HMVP candidates is maintained. When a new slice is encountered, the table is emptied. Whenever there is an inter-coded block, the associated motion information is added to the last entry of the table as a new HMVP candidate. The overall coding flow is shown in Figure 30.

In one example, the table size is set to L (e.g., L=16 or 6 or 44), indicating that up to L HMVP candidates can be added to the table.

In one embodiment (corresponding to Example 11.g.i), if there are more than L HMVP candidates from previously coded blocks, a First-In-First-Out (FIFO) rule is applied so that the table always contains the most recent L previously coded motion candidates. Figure 31 shows an example of applying the FIFO rule to remove HMVP candidates and add new ones to the table used in the proposed method.

In another embodiment (corresponding to invention 11.g.iii), whenever a new motion candidate is added (e.g., the current block is inter-coded and in non-affine mode), we first apply a redundancy check process to identify whether there is a same or similar motion candidate in the LUT.

Some examples are shown below:

Figure 32A shows an example where the LUT is full before adding new motion candidates.

Figure 32B shows an example where the LUT was not full before adding new motion candidates.

Figures 32A and 32B both show an example of a LUT update method based on redundancy elimination (removing one redundant motion candidate).

Figures 33A and 33B show two example implementations of the redundancy removal based LUT update method (removing multiple redundant motion candidates, two candidates are shown in the figure).

Figure 33A shows an example where the LUT is full before adding new motion candidates.

Figure 33B shows an example where the LUT was not full before adding new motion candidates.

HMVP candidates may be used in the merge candidate list construction process. After the TMVP candidate, insert all HMVP candidates from the last entry in the table to the first entry (or the last K0 HMVP, e.g., K0=16 or 6). Apply pruning to the HMVP candidates. End the merge candidate list construction process when the total number of available merge candidates reaches the signaled maximum allowed merge candidates. Alternatively, end the extraction of motion candidates from the LUT when the total number of added motion candidates reaches a given value.

Similarly, HMVP candidates may be used in the AMVP candidate list construction process. Insert the motion vectors of the last K1 HMVP candidates in the table after the TMVP candidates. Construct the AMVP candidate list using only HMVP candidates that have the same reference picture as the AMVP target reference picture. Apply pruning to the HMVP candidates. In one example, K1 is set to 4.

28 is a block diagram of a video processing device 2800. The device 2800 may be used to implement one or more of the methods described herein. The device 2800 may be implemented by a smartphone, a tablet, a computer, an Internet of Things (IoT) receiver, etc. The device 2800 may include one or more processors 2802, one or more memories 2804, and video processing hardware 2806. The one or more processors 2802 may be configured to implement one or more of the methods described herein. The one or more memories 2804 may be used to store data and codes used to implement the methods and techniques described herein. The video processing hardware 2806 may be used to implement the techniques described herein in a hardware circuit.

29 is a flowchart of an example video decoding method 2900. The method 2900 includes maintaining (2902) tables, each table including a set of motion candidates, each motion candidate associated with corresponding motion information. The method 2900 further includes converting (2904) between the first video block and a bitstream representation of the video including the first video block, the converting including processing the motion information of the first video block using at least a portion of the set of motion candidates as predictors.

For method 2900, in some embodiments, the motion information includes at least one of a prediction direction, a reference picture index, a motion vector value, an intensity compensation flag, an affine flag, a motion vector difference precision, or a motion vector difference value. Additionally, the motion information may further include block location information to indicate where the motion information comes from. In some embodiments, the video block is a CU or a PU, and the portion of the video may correspond to one or more video slices or one or more video pictures.

In some embodiments, each LUT includes an associated counter, which is initialized to a zero value at the beginning of the portion of the video and is incremented for each coded video region in the portion of the video. The video region includes one of a coding tree unit, a coding tree block, a coding unit, a coding block, or a prediction unit. In some embodiments, the counter indicates, for a corresponding LUT, the number of motion candidates that have been removed from the corresponding LUT. In some embodiments, the set of motion candidates may have the same size for all LUTs. In some embodiments, the portion of the video corresponds to a video slice, and the number of LUTs is equal to N*P, where N is an integer representing the number of LUTs per decoding thread, and P is an integer representing the number of rows or tiles of the largest coding unit in a video slice. Further details of method 2900 are described in the examples provided in Chapter 4 and in the examples listed below.

Features and embodiments of the above-mentioned methods/techniques are described below.

1. A video processing method comprising: maintaining tables, each table including a set of motion candidates, each motion candidate being associated with corresponding motion information; and converting between a first video block and a bitstream representation of a video including the first video block, the converting including processing the motion information of the first video block using at least a portion of the set of motion candidates as predictors.

2. The method of claim 1, wherein the table includes motion candidates derived from previously decoded video blocks that were decoded before the first video block.

3. The method of claim 1, wherein performing the transformation includes performing an Advanced Motion Vector Prediction (AMVP) candidate list derivation process using at least a portion of the set of motion candidates.

4. The method of claim 3, wherein the AMVP candidate list derivation process includes checking motion candidates from one or more tables.

5. The method according to any one of claims 1 to 4, wherein performing the transformation includes checking motion candidates, and a motion vector associated with the checked motion candidate is used as a motion vector predictor for encoding the motion vector of the first video block.

6. The method of claim 4, wherein the motion vector associated with the checked motion candidate is added to the AMVP motion candidate list.

7. The method of claim 1, wherein performing the transformation includes checking at least a portion of the motion candidates based on a rule.

8. The method of claim 7, wherein the rule enables the check if the AMVP candidate list is not full after checking the TMVP (Temporal Motion Vector Prediction) candidates.

9. The method of claim 7, wherein the rule enables the check if the AMVP candidate list is not full after selecting and pruning from spatial neighborhoods and before inserting the TMVP candidate.

10. The method of claim 7, wherein the rule enables the check if i) no AMVP candidates from the upper neighboring block exist without scaling, or ii) no AMVP candidates from the left neighboring block exist without scaling.

11. The method of claim 7, wherein the rule enables the check if pruning is applied before adding a motion candidate to the AMVP candidate list.

12. The method of claim 1, checking for motion candidates that have the same reference picture as the current reference picture.

13. The method of claim 12, further checking motion candidates that have a different reference picture than the current reference picture.

14. The method of claim 13, wherein checking motion candidates with the same reference picture is performed prior to checking motion candidates with different reference pictures.

15. The method of claim 1, further comprising an AMVP candidate list construction process including a pruning step prior to adding motion vectors from the motion candidates to the table.

16. The method of claim 15, wherein the pruning step includes comparing the motion candidates with at least a portion of the available motion candidates in an AMVP candidate list.

17. The method of claim 15, wherein the pruning step includes multiple steps, the number of which is a function of the number of spatial or temporal AMVP candidates.

18. The method of claim 17, wherein the steps are such that if M candidates are available in the AMVP candidate list, the pruning is applied to only K AMVP candidates, where K≦M and K and M are integers.

19. The method according to claim 1, wherein the conversion includes performing Symmetric Motion Vector Difference (SMVD) processing using differences between several motion vectors.

20. The method according to claim 1, wherein the conversion includes performing SMV (Symmetric Motion Vector) processing using several motion vectors.

21. The method of claim 7, wherein the rule enables the check if the AMVP candidate list is not full after inserting an AMVP candidate.

22. The method of claim 1, further comprising enabling checking of motion candidates in the table, the checking being enabled before checking other candidates derived from spatial or temporal blocks, the other candidates including AMVP candidates, SMVD candidates, SMV candidates, or affine inter candidates.

23. The method of claim 1, further comprising enabling a check of motion candidates in a table, the check being enabled if there is at least one motion candidate in the table.

24. The method of claim 1, in which, for a motion candidate that is a bidirectional prediction candidate, a reference picture in a first reference picture list is checked before a reference picture in a second reference picture list is checked, and the first reference picture list is a current reference target picture list.

25. The method according to claim 1 or 2, in which, for a motion candidate that is a bidirectional prediction candidate, a reference picture in a first reference picture list is checked before a reference picture in a second reference picture list is checked, and the second reference picture list is a current reference target picture list.

26. The method of claim 1, wherein a reference picture in a first reference picture list is checked before a reference picture in a second reference picture list.

27. The method of claim 1, wherein performing the conversion includes generating the bitstream representation from the first video block.

28. The method of claim 1, wherein performing the conversion includes generating the first video block from a bitstream representation.

29. The method according to any one of claims 1 to 28, wherein the motion candidates are associated with motion information including at least one of a prediction direction, a reference picture index, a motion vector value, an intensity compensation flag, an affine flag, a motion vector difference precision, or a motion vector differential value.

30. The method of any one of claims 1 to 29, wherein one motion candidate is associated with an intra prediction mode used for an intra-coded block.

31. A method according to any one of claims 1 to 29, in which one motion candidate is associated with multiple IC (Illumination Compensation) parameters used for an IC-coded block.

32. A method according to any one of claims 1 to 29, in which one motion candidate is associated with a filter parameter used in the filtering process.

33. The method of any one of claims 1 to 29, further comprising updating one or more tables based on the transformation.

34. The method of claim 33, wherein updating one or more tables includes updating one or more tables based on motion information of the first video block after performing the transformation.

35. The method of claim 34, further comprising converting between subsequent video blocks of the video and a bitstream representation of the video based on the updated table.

36. An apparatus comprising a processor and a non-transitory memory having instructions stored therein, the instructions, when executed by the processor, causing the processor to perform a method according to any one of paragraphs 1 to 35.

37. A computer program product stored on a non-transitory computer-readable medium, comprising program code for executing the method according to any one of claims 1 to 35.

While specific embodiments of the disclosed technology have been described above for purposes of illustration, it will be understood that various modifications may be made without departing from the scope of the invention. Accordingly, the disclosed technology is not to be limited except as by the appended claims.

Implementations of the disclosed and other embodiments, modules, and functional operations described herein, including the structures disclosed herein and their structural equivalents, may be implemented in digital electronic circuitry, or computer software, firmware, or hardware, or in one or more combinations thereof. The disclosed and other embodiments may be implemented as one or more computer program products, i.e., one or more modules of computer program instructions encoded on a computer-readable medium for implementation by or for controlling the operation of a data processing apparatus. The computer-readable medium may be a machine-readable storage device, a machine-readable storage substrate, a storage device, a composition of matter that provides a machine-readable propagated signal, or one or more combinations thereof. The term "data processing apparatus" includes all apparatus, devices, and machines for processing data, including, for example, a programmable processor, a computer, or multiple processors, or computers. In addition to hardware, the apparatus may include code that creates an environment for the execution of the computer program, such as code that constitutes a processor firmware, a protocol stack, a database management system, an operating system, or one or more combinations thereof. A propagated signal is an artificially generated signal, for example a machine-generated electrical, optical, or electromagnetic signal, that is generated to encode information for transmission to an appropriate receiving device.

A computer program (also called a program, software, software application, script, or code) can be written in any form of programming language, including compiled or interpreted languages, and it can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file in a file system. A program may be recorded as part of a file that holds other programs or data (e.g., one or more scripts stored in a markup language document), may be stored in a single file dedicated to the program, or may be stored in multiple coordinating files (e.g., files that store one or more modules, subprograms, or portions of code). A computer program can be deployed to run on one computer located at one site, or on multiple computers distributed across multiple sites and interconnected by a communications network.

The processes and logic flows described herein may be performed by one or more programmable processors executing one or more computer programs to perform functions by operating on input data and generating output. The processes and logic flows may also be performed by, and devices may also be implemented as, special purpose logic circuits, such as Field Programmable Gate Arrays (FPGAs) or Application Specific Integrated Circuits (ASICs).

Processors suitable for executing computer programs include, for example, both general purpose and special purpose microprocessors, as well as any one or more processors of any kind of digital computer. Typically, a processor receives instructions and data from a read-only memory or a random access memory or both. The essential elements of a computer are a processor for executing instructions and one or more storage devices for storing instructions and data. Typically, a computer may include one or more mass storage devices, e.g., magnetic, magneto-optical, or optical disks, for storing data, or may be operatively coupled to receive data from or transfer data to these mass storage devices. However, a computer need not have such devices. Computer-readable media suitable for storing computer program instructions and data include all forms of non-volatile memory, media, and storage devices, including, for example, EPROM, EEPROM, flash storage devices, magnetic disks, e.g., internal hard disks or removable disks, magneto-optical disks, and semiconductor storage devices such as CD-ROM and DVD-ROM disks. The processor and memory may be supplemented by, or incorporated in, special purpose logic circuitry.

While this patent specification contains many details, these should not be construed as limiting the scope of any invention or the scope of the claims, but rather as descriptions of features that may be specific to particular embodiments of a particular invention. Certain features described in this patent specification in the context of another embodiment may be implemented in combination in one example. Conversely, various features described in the context of a single example may be implemented in multiple embodiments separately or in any suitable subcombination. Furthermore, although features may be described above as acting in a particular combination and initially claimed as such, one or more features from a claimed combination may, in some cases, be extracted from the combination, and the claimed combination may be directed to a subcombination or a variation of the subcombination.

Similarly, although operations are shown in a particular order in the figures, this should not be understood as requiring that such operations be performed in the particular order or sequential order shown, or that all of the operations shown be performed, to achieve desired results. Additionally, the separation of various system components in the examples described in this patent specification should not be understood as requiring such separation in all embodiments.

Only some implementations and examples have been described, and other embodiments, extensions and variations are possible based on what is described and illustrated in this patent specification.

Claims (26)

constructing a motion candidate list for a current video block, during said constructing, at least one motion candidate in a table is selectively checked in order after checking at least one block each having a predefined relative position with respect to the current video block;
determining motion information for the current video block using the motion candidate list;
coding the current video block based on the determined motion information; and
having
the table includes one or more motion candidates derived from one or more previously coded video blocks that were coded before the current video block;
A method of video processing, wherein an arrangement of the motion candidates in the table is based on an order of addition of the motion candidates to the table. The method of claim 1, wherein the at least one block comprises a temporal block in a different picture than the picture that contains the current video block. The method of claim 1 or 2, wherein the at least one block comprises a given spatial block located in the same picture as the picture containing the current video block. the at least one block includes a last spatial block of a plurality of spatial blocks checked in sequence;
The method of claim 1 , wherein the spatial blocks are located in the same picture as a picture containing the current video block. The method of any one of claims 1 to 4, wherein the motion candidate list is a motion vector prediction list. The method of claim 5, further comprising adding a motion vector of the at least one motion candidate of the table to the motion vector prediction list. The method of claim 5 or 6, wherein checking the at least one motion candidate in the table includes checking reference picture information of the motion candidate. The method of any one of claims 1 to 4, wherein the motion candidate list is a merge candidate list. The method of claim 8, further comprising adding the at least one motion candidate from the table to the merge candidate list. The method of claim 9, wherein the at least one motion candidate in the table is checked before checking other merge candidates associated with the spatial or temporal block. The method according to any one of claims 1 to 10, wherein the motion vector of the at least one motion candidate of the table or the at least one motion candidate is added to the motion candidate list after all of the at least one motion candidate derived from the at least one block. The method of any one of claims 1 to 11, wherein the predefined relative position of the at least one block is different for different objects to be coded. If at least one condition is met, at least one motion candidate in the table is checked;
13. The method of claim 1, wherein the at least one condition is based on a current number of motion candidates in the motion candidates table after checking the at least one block or a current number of motion candidates in the table. The method of claim 13, wherein the at least one condition further includes the current number of motion candidates in the motion candidate list not reaching a particular value. The method of claim 14, wherein the particular value is associated with a maximum allowable number of motion candidates in the motion candidate list. The method of any one of claims 13 to 15, wherein the at least one condition includes the current number of motion candidates in the table being greater than zero. The method of any one of claims 1 to 16, wherein the motion candidates of the table are associated with motion information including at least one of a prediction direction, a reference picture index, a motion vector value, an intensity compensation flag, an affine flag, a motion vector differential precision, a motion vector differential value, or a filtering parameter. The method of any one of claims 1 to 17, further comprising updating the table based on the motion information of the current block. The method of any one of claims 1 to 18, wherein one or more motion candidates in the table are deleted due to adding a new motion candidate to the table if the table is full before the new motion candidate is added to the table. The method of any one of claims 1 to 19, wherein the current block is coded in a non-affine inter mode or an affine inter mode. 21. The method of claim 1, further comprising adding a motion vector of the at least one motion candidate of the table to the motion candidate list based on a result of the checking. The method of any one of claims 1 to 21, wherein the coding includes decoding the current video block from the video block. The method of any one of claims 1 to 21, wherein the coding comprises encoding the current video block into a bitstream representation. 1. An apparatus for coding video data having a processor and a non-transitory memory having instructions, comprising:
The instructions, when executed by the processor, cause the processor to:
constructing a motion candidate list for a current video block of a video, during said constructing, at least one motion candidate in a table is selectively checked in order after checking at least one block each having a predefined relative position with respect to the current video block;
determining motion information for the current video block using the motion candidate list;
coding the current video block based on the determined motion information; and
Run the command,
the table includes one or more motion candidates derived from one or more previously coded video blocks that were coded before the current video block;
An apparatus, wherein an ordering of the motion candidates in the table is based on an order of addition of the motion candidates to the table. The processor:
constructing a motion candidate list for a current video block of a video, during said constructing, at least one motion candidate in a table is selectively checked in order after checking at least one block each having a predefined relative position with respect to the current video block;
determining motion information for the current video block using the motion candidate list;
coding the current video block based on the determined motion information; and
Run the command,
the table includes one or more motion candidates derived from one or more previously coded video blocks that were coded before the current video block;
A non-transitory computer-readable storage medium having stored thereon instructions, wherein an ordering of the motion candidates in the table is based on an order of addition of the motion candidates to the table. 1. A method for storing a video bitstream, comprising:
constructing a motion candidate list for a current video block of the image, during said construction at least one motion candidate in a table being checked selectively in sequence after checking at least one block each having a predefined relative position with respect to the current video block;
determining motion information for the current video block using the motion candidate list;
encoding the current video block into the bitstream based on the determined motion information; and
storing the bitstream on a non- transitory computer-readable recording medium;
the table includes one or more motion candidates derived from one or more previously coded video blocks that were coded before the current video block;
A method according to claim 1, wherein an ordering of the motion candidates in the table is based on an order of addition of the motion candidates to the table.